The long-term objective of this research is to understand how the brain controls voluntary movements in order to treat disorders of the motor-output portion of the nervous system. Specifically, this proposal investigates how a particular cortical region - dorsal premotor cortex (PMd) - is involved in controlling reaching, from the initial trajectory to the final position. Before movement has been initiated, properties of an upcoming reach are predicted by neural activity in PMd. For this reason, PMd is thought to be involved in planning reaching movements. In real-life situations, there are many different trajectories which result in the correct hand position at the end of a reach. Exactly which trajectory is selected depends on many factors, such as constraints in the environment or internally-generated motives. Depending on the circumstances, the trajectory of the limb may be as important as the final position in achieving the desired goal. By contrast, previous studies in non-human primates have overwhelmingly utilized straight-line movements from one point to another. Movement direction and target position are highly correlated in this case, which confounds efforts to determine how these different components of the reach are precisely represented by the brain. To accurately study how different features of a movement modulate neural activity, it is necessary to dissociate trajectory from target position. The research proposed here accomplishes this by placing obstacles in a virtual environment, inducing subjects to choose curved reaching trajectories. Target and obstacle positions are varied from reach to reach, requiring subjects to develop a new plan before each movement. Subjects' hand and eye positions are recorded concurrently with electrical activity of individual neurons in PMd during the planning and execution of the reach. By examining the relationship between neural activity and motor output under these conditions, the role of PMd in generating voluntary reaching movements will be better understood.